Feature Request: Clarify age group stratification

[ablack3]

Is your feature request related to a problem? Please describe.

I found the age group specification a little unclear in the export function. Here is the documentation.

The age groups are specified with breaks such as c(0, 18, 150)
Then age groups in the results are given as "0 to 18", "18 to 150"
It's not so clear then which age groups the 18 year olds are included in. Based on the documentation I'd assume 18 year olds are included in "0 to 18" but not "18 to 150". So it seems the lower bound is not included in the interval.

However if I were to specify c(18, 65, 150) then I'd get age groups "18 to 65", "65 to 150" and the 18 year olds would be included in the interval "18 to 65" (lower bound is inclusive).

This is in contrast to other packages we use for Darwin in which the bounds of the intervals are included in the interval, e.g. 18 to 150 always means that 18 year olds are included in the interval. Some packages like SummariseCharacteristics require that the intervals are not overlapping while others like IncidencePrevalence will allow overlapping intervals.

Describe the solution you'd like
I would suggest that the labels of the age groups in the output to be consistent, preferably with the darwin packages, where both bounds of the interval are always included in the interval.

For example:
The age groups are specified with breaks such as c(0, 18, 150)
Then age groups in the results would be given as "0 to 17", "18 to 150" making it clear which group the 18 year olds are included in.

[mvankessel-EMC]

Thanks @ablack3 I'd like to move away from the way how strata are defined right now. And move to a more IncidencePrevalence-esk way of doing things, where you can add an arbitrary column, and stratify by that (#273).